The Safety of Laparoscopic Sleeve Gastrectomy in

OBES SURG DOI 10.1007/s11695-015-1590-1

ORIGINAL CONTRIBUTIONS

The Safety of Laparoscopic Sleeve Gastrectomy in Patients Receiving Chronic Anticoagulation Therapy: A Case-Matched Study Olivier Gerin & Lionel Rebibo & Abdennaceur Dhahri & Jean-Marc Regimbeau

# Springer Science+Business Media New York 2015

Abstract Background Obesity is linked to cardiac disorders with a relative risk of atrial fibrillation of 1.5 (requiring the use of chronic anticoagulation therapy, CAT). However, CAT is a known risk factor for postoperative bleeding after elective surgery. The primary objective of the present study was to evaluate the short- and long-term complications of laparoscopic sleeve gastrectomy (LSG) in patients receiving CAT. Methods This is a retrospective analysis of a prospective database of CAT patients undergoing LSG between March 2004 and December 2012. This LSG-CAT group was matched 1:2 on preoperative data with patients not receiving CAT (LSGcontrol group). Primary efficacy criterion was the frequency of CAT-related complications. Secondary efficacy criteria were the major postoperative complications, frequency of revisional surgery, long-term CAT-related complications, and a change in the dose level of oral anticoagulants. Results The LSG-CAT group consisted of 15 patients with a median age of 54 years (32–65). The LSG-control group consisted of 30 patients. Median operating time was 75 min in both groups (p=0.33). Major complication rates in the LSG-CAT and LSG-control groups were 13.3 and 3.3 %, respectively (p=0.20), with one case of postoperative bleeding in each group (6.7 and 3.3 %, p=0.6); incidence of revisional surgery was 13.3 and 3.3 % (p=0.2). There were no postoperative mortalities. After a median follow-up of 14 months (9– 43), no changes in the dose level of oral anticoagulants were reported.

O. Gerin : L. Rebibo : A. Dhahri : J.50 kg/m2 5 (33.3) Hypertension (%) 8 (53.3) Diabetes mellitus (%) 3 (20) Dyslipidemia (%) 8 (53.3) OSA (%) 7 (46.7) Intra- and postoperative data Operating time (min) 75 (55–135) Conversion (%) 0 (0) Length of hospital stay 3 (3–11) (days) Mortality (%) 0 (0) Complications (%) 4 (26.7) Clavien grade ≥3 2 (13.3) Revisional surgery (%) 2 (13.3) Gastric fistula (%) 1 (6.7) Hemorrhage (%) 1 (6.7)

LSG-control group (n=30)

p value

54.5 (25–59) 30 47.2 (38.2–62.7) 8 (26.7) 15 (50) 7 (23.3) 15 (50) 15 (50)

0.89 0.97 1.0 0.83 0.79 0.67 0.83

75 (45–150) 2 (6.7) 3 (1–8)

0.33 0.30 0.52

0 (0) 4 (13.3) 1 (3.3) 1 (3.3) 0 (0) 1 (3.3)

0.29 0.20 0.20 0.15 0.60

BMI body mass index, OSA obstructive sleep apnea

laparotomy in the LSG-control group (6.7 %) due to adhesions (p=0.306). The median total operating time was 75 min (50–135) in the LSG-CAT group and 75 min (45–150) in the LSG-control group (p=0.333) (Table 1). There were no intraoperative bleeding events in both groups. After 2009, abdominal drainage was not implemented in any of the patients in the LSGCAT group. Postoperative Data Mortality and Morbidity There were no postoperative deaths in either group. Four postoperative complications (26.6 %) occurred in the LSG-CAT group, including two major complications (13.3 %) requiring revisional surgical procedures for gastric fistula and postoperative bleeding. The major complications included a gastric fistula (6.7 %) and a postoperative hematoma (6.7 %) in the LSG-CAT group (Fig. 1). The median length of hospitalization in the LSG-CAT group was 3 days (3–11). In the LSG-control group, there were four postoperative complications (13.3 %), including one major

complication requiring revisional surgery for postoperative bleeding (3.3 %). The three other postoperative complications were ischemia of the upper pole of the spleen (requiring the extension of hospitalization and treatment with analgesics, including morphine), one case of acute coronary syndrome (treated with medication only), and one case of hypertension. There were no cases of postoperative gastric fistula. The median length of hospitalization in the LSG group was 3 days (1–8). The LSG-CAT and LSG-control groups did not differ significantly in terms of the study’s primary endpoint, i.e., the complications rate (26.6 vs. 13.3 %, respectively, p=0.294). This was also true for serious complications (Clavien score ≥3), with frequencies of 13.3 and 3.3 %, respectively (p= 0.152) (Table 1). Weight Loss In the LSG-CAT group, the median BMI was 37 kg/m2 (29.4– 54.4) at 6 months and 31 kg/m2 (21.6–40.3) at 12 months. In the LSG-control group, the median BMI was 38 kg/m2 (22.8– 47.9) at 6 months and 33.7 kg/m2 (23.7–40.8) at 12 months. The median EWL at 6 months was 46 % (15.5–88) in the LSG-CAT group and 44 % (14.6–91.8) in the LSG-control group (p=0.784). The median EWL at 12 months was 64 % (42–124) in the LSG-CAT group and 52 % (31.6–133) in the LSG-control group (p=0.125). The percentage weight loss at 6 months in the LSG-CAT group and LSG-control group was 21 % (9–31) and 19 % (8.4–43.8), respectively (p=0.916). The percentage weight loss at 12 months in the LSG-CAT group and LSG-control group was 31 % (16–47) and 26.6 (15–50), respectively (p= 0.205) (Table 2). Complications Related to CAT Chronic anticoagulation therapy was reintroduced by the patient’s primary care physician after a median time interval of 15 days (14–19). The median follow-up period was 14.5 months (8–30). During follow-up, there were no hemorrhages related to excessive CAT and no thrombotic events related to insufficient CAT. After 14.5 months of follow-up, none of the patients had required a change in the dose of oral anticoagulants.

Discussion The surgical management of obese patients receiving CAT is a real problem. Elective surgery requires a switch to LMWH or intravenous unfractionated heparin in case of elective surgery, whereas prothrombin complex concentrate can be used during emergency surgery. The results of several series have shown

OBES SURG Fig. 1 An abdominal CT scan showing a perigastric hematoma (white arrow) requiring revisional surgery

that patients with preoperative CAT have an increased risk of postoperative complications, such as bleeding and thrombosis [20]. Bariatric surgery dates from the 1960s and was first performed by laparotomy [21]. Despite good results in terms of weight loss, bariatric surgery used to be associated with significant morbidity (especially concerning the abdominal wall). In the 1990s, the rise of laparoscopy and the development of banding [22] as minimally invasive surgery prompted the further development of bariatric surgery. Laparoscopic surgery is of value in reducing postoperative morbidity in colorectal [23] and liver surgery [24]. It also helps to reduce pulmonary complications via greater postoperative pulmonary mobility [25]. Laparoscopy also had an impact on gastric bypass surgery [26, 27] because it has enabled a reduction in the length of hospital stay and in blood loss (often a risk factor for postoperative complications) [28].

Table 2 Changes in BMI, EWL, and percentage weight loss in the LSG-CAT and LSG-control groups over 12 months of follow-up LSG-CAT group (n=15) Initial BMI BMI at 6 months EWL at 6 months Percentage weight loss at 6 months BMI at 12 months EWL at 12 months Percentage weight loss at 12 months

LSG-control group (n=30)

p value

47.7 (37–63) 47.2 (33.9–54.3) 0.97 37 (26.6–54.4) 38 (23.3–52.2) 0.93 46 (15.5–88) 44 (14–91.8) 0.78 21 (9–29) 19 (8.4–35) 0.92 31 (21.6–40.3) 64 (42–124) 31 (16–47)

BMI body mass index, EWL excess weight loss

34 (20.1–48) 0.07 52 (31.6–133) 0.12 26 (15–49.5) 0.20

Logically, the indications of bariatric surgery have been extended to more vulnerable patients (such as patients receiving CAT or with liver cirrhosis) for whom the benefits of surgery are nevertheless likely to outweigh the risks [29]. The few series to have evaluated the outcomes of bariatric surgery in patients with CAT all looked at RYGBP. Mourelo et al. [12] studied the frequency of hemorrhage and thromboembolic complications after RYGBP in 1700 patients. Twenty-one of the patients were receiving CAT. There were no conversions to laparotomy and no intraoperative bleeding events. Three patients (14 %) had postoperative hemorrhage, but only one required revisional surgery. However, a limitation of the study of Mourelo et al. was its purely descriptive nature; patients receiving CAT were not compared with the other patients. Bechtel et al. [13] studied the 30-day readmission rate and the causes of readmission in 1338 patients having underdone RYGBP. There were 133 readmissions, including 22 for hemorrhage. Of these 22 patients, 6 had been receiving CAT (with warfarin). These six patients accounted for 15.8 % of all patients receiving CAT (n = 38) and more than 50 % of readmitted patients with CAT (n=11). In the control population, 15 patients (1.1 %) were readmitted for hemorrhage. The readmission rate was significantly higher in patients with CAT (p